JAKARTA - Apple is facing a serious lawsuit in the United States after being accused of using videos from YouTube without permission to train their artificial intelligence (AI) models.
The lawsuit was filed on April 3, 2026 in a California federal court, on the basis of violation of the Digital Millennium Copyright Act (DMCA). A number of YouTube content creators, including Ted Entertainment - which manages the h3h3Productions channel - are the main parties in this class action.
In addition to Ted Entertainment, two other channels, namely MrShortGameGolf and Golfholics, also joined the lawsuit. They accused Apple of systematically downloading and archiving videos using a method that allegedly violated the platform protection system.
In court documents, Apple is said to have used scraping techniques with constantly changing IP addresses to avoid detection, then collected the data into a large dataset used to train the "Apple AI Video" model.
As evidence, the plaintiff refers to a dataset called Panda-70M, which is said to consist entirely of scraped YouTube videos. In the dataset, it is recorded that hundreds of videos belonging to the plaintiffs are used, including 438 videos from Ted Entertainment's channel.
The plaintiffs consider this practice as a serious violation of copyright and "an attack on the content creator community" whose work is used without compensation.
"This is an unacceptable attack on creators, where their content is used to support a trillion-dollar AI industry without permission," the lawsuit reads.
In addition to demanding maximum compensation, the plaintiff also asked the court to issue an injunction (injunction) so that Apple no longer uses data obtained illegally, as well as demanding additional compensation and legal costs.
This case adds to the long list of legal conflicts related to public data-based AI training. Previously, companies such as Meta, Nvidia, and ByteDance also faced similar accusations.
Ironically, Apple has long been known to try to take a more "ethical" approach to AI development, including exploring content licensing from publishers such as Conde Nast and NBC News. However, this latest lawsuit shows that even big players find it difficult to avoid controversy in the scramble for data to train AI models.
The case also comes just weeks after Apple was dragged into another lawsuit over its use of the "The Pile" dataset, which allegedly contained copyrighted material collected without permission.
With increasing legal pressure from creators and content owners, the outcome of this case has the potential to become an important precedent in determining the legal limits of using data for future AI training.
Follow VOI Whatsapp Channel
The English, Chinese, Japanese, Arabic, and French versions are automatically generated by the AI. So there may still be inaccuracies in translating, please always see Indonesian as our main language. (system supported by DigitalSiber.id)
